Wind profiles in the troposphere represent one of the major unfilled measurement needs for improving global weather forecasting and analysis. Various approaches for global tropospheric wind profiling that are based on Doppler lidar have been proposed and/or developed. However, these techniques have various limitations, especially when the wavelength of the light in connection with which the shift is being measured is shorter than 5-10 micrometers. Such limitations result from various factors, such as distortion of propagating optical wavefronts (as in the twinkling of stars) by inhomogeneities in the refractive index of the atmosphere, the size of the Doppler shift (e.g., where it is too large for direct frequency measurement and too small for normal spectroscopic techniques), and inefficiency in the use of the received light.
In general, techniques for measuring wind profiles in the troposphere have involved the use of Doppler lidars (light detection and ranging). In such systems, light from a laser source is directed towards a target volume in the atmosphere. Particles and molecules in the target volume will scatter the light, with some of the scattered light being reflected back towards the laser source. By measuring the change in the wavelength of the reflected light as compared to the wavelength of the laser source, the velocity component of particles and molecules within the volume along the direction of the laser beam, can be determined.
In order to determine the Doppler shift, heterodyne detection, in which the optical return is mixed with an optical local oscillator to give the Doppler difference frequency (beat frequency) in the radio frequency (RF) region, has been used. In such systems, the radio frequency signal is typically digitized and processed via digital time delay autocorrelation (single or multiple lag) to give the Doppler shift from the wind component. Alternatively, radio frequency spectrum analyzers based on a scanning filter or a discrete filter bank can be used. However, heterodyne detection has rarely been used at wavelengths shorter than about 2.1 micrometers, because the Doppler difference frequency is too high to digitize and process electronically, and because the heterodyne conversion gain is not needed with low noise detectors at shorter wavelengths.
Direct detection of the Doppler shift is commonly used with visible (VIS) and ultraviolet (UV) Doppler lidars. In principle, signal processing and direct detection devices can be done either in optical frequency (wavelength) space or in time delay (autocorrelation) space. More particularly, direct detection devices have used Fabry-Perot interferometers, which are narrow band, optical, band pass filters. Specific Fabry-Perot approaches have been based on edge detection, a linearized ring pattern, and other techniques. In general, Fabry-Perot approaches are the optical analog of electronic scanning filter or filter bank spectrum analyzers.
In connection with the remote measurement of wind velocities, Doppler lidars using relatively short wavelengths (e.g., near 1 micrometer) are desirable. In particular, shorter wavelength lidar offers better range and velocity resolution for measurements in the turbulent boundary layer because the product of range resolution and spectral width of the velocity spectrum is proportional to the wavelength. In addition, the quantized backscattering coefficient (i.e., the number of photons backscattered for a given transmitted pulse energy) may be greater near 1 micrometer than at longer wavelengths, such as about 10.6 micrometers, for particular size distributions having a relatively small mean particle size. However, operating at shorter wavelengths has been difficult using conventional techniques. For example, in connection with optical heterodyne detection, the beat frequencies at such short wavelengths are high enough (about 2 MHz per 1 meter per second near a wavelength of 1 micrometer) to require expensive analog to digital converters and extremely high data rates for digital processing. Furthermore, inhomogeneities in atmospheric refractive index limit the effective aperture of lidar telescopes for heterodyning.
Doppler lidars operating at wavelengths near 0.5 micrometers that use optical heterodyne detection have been investigated. However, because of difficulties with optical heterodyning in the associated spectral region, they have not been pursued. Instead, Doppler lidars operating at 0.5 micrometers and shorter wavelengths have been based on signal processing using narrow transmission filters (the “edge” technique) or using configurations of Fabry-Perot interferometers, including either scanning, multiple fixed etalons, or displaced ring patterns with a spatially extended scattering source. However, such devices make relatively inefficient use of photons and require relatively complex hardware.